Oncological and functional results were assessed in 47 type C and/or D jugular paraganglioma operated on between 1984 and 1998 using the classical infratemporal fossa type A approach (mean follow-up = 66 months). In 24 instances, however, the facial nerve was not re-routed. Total resection was achieved in 33 cases (70 per cent). In 25 patients available for follow-up this resulted in a 92 per cent cure rate while two patients (eight per cent) developed recurrences that are being followed-up clinically and radiologically. Sub-total resection, leaving infracentimetric tumour remnants after being coagulated, was achieved in 14 cases (30 per cent). In 11 patients available for follow-up, only three cases developed tumour regrowth (27 per cent) that was controlled by salvage irradiation or surgery while in the other cases tumour remnants remained stable (73 per cent). Symptomatic post-operative lower cranial nerve impairment was observed in 23 per cent. When results were analysed depending on whether the facial nerve had been re-routed (n = 18) or not (n = 24), the incidence of facial paralysis HB grade III or more at one year was 33 per cent and eight per cent, respectively. Total resection was achieved in 56 per cent when the facial nerve was re-routed versus 75 per cent when it was not, the difference being due to a higher incidence of large tumours in the first group. The present study suggests that: 1) surgical resection of jugular paraganglioma provides overall satisfactory results, i.e. a 86 per cent rate of either cure or tumour remnant stabilization, but carries a significant risk of iatrogeny; 2) complete tumour removal should not be attempted, especially in patients over 60 years of age with no pre-operative neurological deficits, since leaving infracentimetric tumour remnants has no major detrimental effect on the final outcome; 3) facial nerve transposition carries a significant risk of cosmetic sequelae while it does not provide significant advantages in terms of tumour resection and long-term oncological control.